forked from dclambert/Python-ELM
/
elm_notebook.py
218 lines (157 loc) · 6.8 KB
/
elm_notebook.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
# -*- coding: utf-8 -*-
# <nbformat>2</nbformat>
# <codecell>
# Demo python notebook for sklearn elm and random_hidden_layer modules
#
# Author: David C. Lambert [dcl -at- panix -dot- com]
# Copyright(c) 2013
# License: Simple BSD
# <codecell>
from time import time
from math import sqrt
from matplotlib.pyplot import plot, scatter, hist
from numpy.ma import mean, std
from sklearn.cluster import k_means
import numpy as np
from elm import ELMClassifier, ELMRegressor, GenELMClassifier, GenELMRegressor
from random_layer import RandomLayer, MLPRandomLayer, RBFRandomLayer, GRBFRandomLayer
# <codecell>
def make_toy():
x = np.arange(0.25, 20, 0.1)
y = x * np.cos(x) + 0.5 * np.sqrt(x) * np.random.randn(x.shape[0])
x = x.reshape(-1, 1)
y = y.reshape(-1, 1)
return x, y
# <codecell>
def res_dist(x, y, e, n_runs=100, random_state=None):
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.4, random_state=random_state)
test_res = []
train_res = []
start_time = time()
for i in range(n_runs):
e.fit(x_train, y_train)
train_res.append(e.score(x_train, y_train))
test_res.append(e.score(x_test, y_test))
if (i % (n_runs / 10) == 0): print("%d" % i, end=' ')
print("\nTime: %.3f secs" % (time() - start_time))
print("Test Min: %.3f Mean: %.3f Max: %.3f SD: %.3f" % (min(test_res), mean(test_res), max(test_res), std(test_res)))
print("Train Min: %.3f Mean: %.3f Max: %.3f SD: %.3f" % (
min(train_res), mean(train_res), max(train_res), std(train_res)))
return (train_res, test_res)
# <codecell>
from sklearn.cross_validation import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.datasets import load_iris, load_digits, load_diabetes, make_regression
stdsc = StandardScaler()
iris = load_iris()
irx, iry = stdsc.fit_transform(iris.data), iris.target
irx_train, irx_test, iry_train, iry_test = train_test_split(irx, iry, test_size=0.2)
digits = load_digits()
dgx, dgy = stdsc.fit_transform(digits.data / 16.0), digits.target
dgx_train, dgx_test, dgy_train, dgy_test = train_test_split(dgx, dgy, test_size=0.2)
diabetes = load_diabetes()
dbx, dby = stdsc.fit_transform(diabetes.data), diabetes.target
dbx_train, dbx_test, dby_train, dby_test = train_test_split(dbx, dby, test_size=0.2)
mrx, mry = make_regression(n_samples=2000, n_targets=4)
mrx_train, mrx_test, mry_train, mry_test = train_test_split(mrx, mry, test_size=0.2)
xtoy, ytoy = make_toy()
xtoy, ytoy = stdsc.fit_transform(xtoy), stdsc.fit_transform(ytoy)
xtoy_train, xtoy_test, ytoy_train, ytoy_test = train_test_split(xtoy, ytoy, test_size=0.2)
plot(xtoy, ytoy)
# <codecell>
# RBFRandomLayer tests
for af in RandomLayer.activation_func_names():
print(af, end=' ')
elmc = ELMClassifier(activation_func=af)
tr, ts = res_dist(irx, iry, elmc, n_runs=200, random_state=0)
# <codecell>
elmc.classes_
# <codecell>
for af in RandomLayer.activation_func_names():
print(af)
elmc = ELMClassifier(activation_func=af, random_state=0)
tr, ts = res_dist(dgx, dgy, elmc, n_runs=100, random_state=0)
# <codecell>
elmc = ELMClassifier(n_hidden=500, activation_func='multiquadric')
tr, ts = res_dist(dgx, dgy, elmc, n_runs=100, random_state=0)
scatter(tr, ts, alpha=0.1, marker='D', c='r')
# <codecell>
elmr = ELMRegressor(random_state=0, activation_func='gaussian', alpha=0.0)
elmr.fit(xtoy_train, ytoy_train)
print(elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test))
plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
# <codecell>
from sklearn import pipeline
from sklearn.linear_model import LinearRegression
elmr = pipeline.Pipeline([('rhl', RandomLayer(random_state=0, activation_func='multiquadric')),
('lr', LinearRegression(fit_intercept=False))])
elmr.fit(xtoy_train, ytoy_train)
print(elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test))
plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
# <codecell>
rhl = RandomLayer(n_hidden=200, alpha=1.0)
elmr = GenELMRegressor(hidden_layer=rhl)
tr, ts = res_dist(mrx, mry, elmr, n_runs=200, random_state=0)
scatter(tr, ts, alpha=0.1, marker='D', c='r')
# <codecell>
rhl = RBFRandomLayer(n_hidden=15, rbf_width=0.8)
elmr = GenELMRegressor(hidden_layer=rhl)
elmr.fit(xtoy_train, ytoy_train)
print(elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test))
plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
# <codecell>
nh = 15
(ctrs, _, _) = k_means(xtoy_train, nh)
unit_rs = np.ones(nh)
#rhl = RBFRandomLayer(n_hidden=nh, activation_func='inv_multiquadric')
#rhl = RBFRandomLayer(n_hidden=nh, centers=ctrs, radii=unit_rs)
rhl = GRBFRandomLayer(n_hidden=nh, grbf_lambda=.0001, centers=ctrs)
elmr = GenELMRegressor(hidden_layer=rhl)
elmr.fit(xtoy_train, ytoy_train)
print(elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test))
plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
# <codecell>
rbf_rhl = RBFRandomLayer(n_hidden=100, random_state=0, rbf_width=0.01)
elmc_rbf = GenELMClassifier(hidden_layer=rbf_rhl)
elmc_rbf.fit(dgx_train, dgy_train)
print(elmc_rbf.score(dgx_train, dgy_train), elmc_rbf.score(dgx_test, dgy_test))
def powtanh_xfer(activations, power=1.0):
return pow(np.tanh(activations), power)
tanh_rhl = MLPRandomLayer(n_hidden=100, activation_func=powtanh_xfer, activation_args={'power':3.0})
elmc_tanh = GenELMClassifier(hidden_layer=tanh_rhl)
elmc_tanh.fit(dgx_train, dgy_train)
print(elmc_tanh.score(dgx_train, dgy_train), elmc_tanh.score(dgx_test, dgy_test))
# <codecell>
rbf_rhl = RBFRandomLayer(n_hidden=100, rbf_width=0.01)
tr, ts = res_dist(dgx, dgy, GenELMClassifier(hidden_layer=rbf_rhl), n_runs=100, random_state=0)
# <codecell>
hist(ts), hist(tr)
print
# <codecell>
from sklearn.svm import SVR
from sklearn.ensemble import RandomForestRegressor
tr, ts = res_dist(dbx, dby, RandomForestRegressor(n_estimators=15), n_runs=100, random_state=0)
hist(tr), hist(ts)
print
rhl = RBFRandomLayer(n_hidden=15, rbf_width=0.1)
tr,ts = res_dist(dbx, dby, GenELMRegressor(rhl), n_runs=100, random_state=0)
hist(tr), hist(ts)
print
# <codecell>
elmc = ELMClassifier(n_hidden=1000, activation_func='gaussian', alpha=0.0, random_state=0)
elmc.fit(dgx_train, dgy_train)
print(elmc.score(dgx_train, dgy_train), elmc.score(dgx_test, dgy_test))
# <codecell>
elmc = ELMClassifier(n_hidden=500, activation_func='hardlim', alpha=1.0, random_state=0)
elmc.fit(dgx_train, dgy_train)
print(elmc.score(dgx_train, dgy_train), elmc.score(dgx_test, dgy_test))
# <codecell>
elmr = ELMRegressor(random_state=0)
elmr.fit(xtoy_train, ytoy_train)
print(elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test))
plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
# <codecell>
elmr = ELMRegressor(activation_func='inv_tribas', random_state=0)
elmr.fit(xtoy_train, ytoy_train)
print(elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test))
plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))